- Email: [email protected]
Cost analysis of ongoing care of patients with end-stage renal disease: The impact of dialysis modality and dialysis access
Cost Analysis of Ongoing Care of Patients With End-Stage Renal Disease: The Impact of Dialysis Modality and Dialysis Access Helen Lee, MA, Braden Manns, MD, Ken Taub, MD, William A. Ghali, MD, Stafford Dean, BSc, David Johnson, MD, and Cam Donaldson, PhD ● Background: Care of patients with end-stage renal disease (ESRD) is important and resource intense. To enable ESRD programs to develop strategies for more cost-efficient care, an accurate estimate of the cost of caring for patients with ESRD is needed. Methods: The objective of our study is to develop an updated and accurate itemized description of costs and resources required to treat patients with ESRD on dialysis therapy and contrast differences in resources required for various dialysis modalities. One hundred sixty-six patients who had been on dialysis therapy for longer than 6 months and agreed to enrollment were followed up prospectively for 1 year. Detailed information on baseline patient characteristics, including comorbidity, was collected. Costs considered included those related to outpatient dialysis care, inpatient care, outpatient nondialysis care, and physician claims. We also estimated separately the cost of maintaining the dialysis access. Results: Overall annual cost of care for in-center, satellite, and home/self-care hemodialysis and peritoneal dialysis were US $51,252 (95% confidence interval [CI], 47,680 to 54,824), $42,057 (95% CI, 39,523 to 44,592), $29,961 (95% CI, 21,252 to 38,670), and $26,959 (95% CI, 23,500 to 30,416), respectively (P < 0.001). After adjustment for the effect of other important predictors of cost, such as comorbidity, these differences persisted. Among patients treated with hemodialysis, the cost of vascular access– related care was lower by more than fivefold for patients who began the study period with a functioning native arteriovenous fistula compared with those treated with a permanent catheter or synthetic graft (P < 0.001). Conclusion: To maximize the efficiency with which care is provided to patients with ESRD, dialysis programs should encourage the use of home/self-care hemodialysis and peritoneal dialysis. © 2002 by the National Kidney Foundation, Inc. INDEX WORDS: Kidney failure; dialysis; costs; economic evaluation; hemodialysis (HD); peritoneal dialysis (PD); arteriovenous fistula (AVF); synthetic graft.
C
ARE OF PATIENTS with end-stage renal disease (ESRD) is resource intense, but also important because renal replacement therapy is life saving. It has been estimated that 2% to 3% of a developed country’s health care budget is spent caring for patients with ESRD, although only 0.07% of the population has ESRD.1-4 Because of high costs and the increasing prevalence of ESRD,1,5 it is important for ESRD programs to develop strategies for more cost-efficient care. The largest area of cost for ESRD programs relates to the provision of outpatient renal replacement therapy.6 There are different methods of renal replacement therapy, including renal transplantation, hemodialysis, and peritoneal dialysis. When determining which modality to encourage, it is important to consider the effect of each modality on survival, health-related quality of life (HRQOL), and resource use. Renal transplantation is associated with superior clinical outcomes, HRQOL, and cost savings compared with dialysis.7,8 Unfortunately, many patients are not candidates for renal transplantation or must wait for a transplant on dialysis therapy.1 Regarding dialytic therapies, survival appears to be similar between peritoneal dialysis and hemodialysis,9-12 and studies suggest that HRQOL is similar for
patients on in-center hemodialysis and peritoneal dialysis therapy.13 Patients treated with home/selfcare hemodialysis may have a slightly greater HRQOL.14,15 Although some patients may have a medical or social preference for one modality, a large proportion of patients starting dialysis therapy could be treated successfully with any of the modalities.16-18 For this group of patients, who From the Departments of Economics, Community Health Sciences, and Medicine, University of Calgary; Alberta Centre of Health Services Utilization Research, Edmonton, Alberta; and the Department of Medicine, Anaesthesiology, Community Health and Epidemiology, University of Saskatchewan, Saskatchewan, Canada. Received February 28, 2002; accepted in revised form April 24, 2002. Work performed at Foothills Medical Centre, Calgary, Alberta, Canada. Supported in part by The Alberta Heritage Foundation for Medical Research and The Center for Advancement of Health, Calgary Regional Health Authority, Calgary, Alberta, Canada. Address reprint requests to Braden Manns, MD, Foothills Medical Centre, 1403 29th St NW, Calgary, Alberta, Canada, T2N 2T9. E-mail: [email protected] © 2002 by the National Kidney Foundation, Inc. 0272-6386/02/4003-0022$35.00/0 doi:10.1053/ajkd.2002.34924
American Journal of Kidney Diseases, Vol 40, No 3 (September), 2002: pp 611-622
611
612
LEE ET AL
may have limited knowledge about kidney failure and its treatment, it is important to educate patients about which modality would be optimal for them, keeping in mind the overall resources of the program and cost of each modality. Previous cost studies are largely outdated19-22; have used aggregate costs or prices23; have been performed outside of North America, where practice patterns may be different24,25; or have studied dialytic modalities (intermittent peritoneal dialysis) not in use today.6 The primary objective of our study is to provide ESRD programs with an updated and accurate itemized list of costs and resources required to treat patients with ESRD undergoing dialysis and contrast differences in resources required for various dialysis modalities. Our secondary objective is to estimate separately the cost of maintaining the dialysis access (arteriovenous fistula [AVF], synthetic graft, or permanent catheter). METHODS
Setting The Conjoint Medical and Research Ethics Board at the University of Calgary (Alberta, Canada) approved the study protocol. In July 1999, a total of 416 patients were treated locally with dialysis (266 patients, in-center hemodialysis; 55 patients, satellite hemodialysis; 79 patients, peritoneal dialysis; 13 patients, home hemodialysis; and 3 patients, self-care hemodialysis). Three hundred thirty-two of these patients had been on dialysis therapy longer than 6 months and were eligible for enrollment. Six months was chosen because (1) by that time, dialysis modality and a permanent vascular access generally have been established, and (2) the goal of this study is to determine the cost of ongoing dialysis, rather than costs associated with initiating dialysis therapy. Patients were approached for enrollment in a consecutive fashion from a randomly generated list of patients treated with each modality. Within the Southern Alberta Renal Program (SARP), which cares for all patients with ESRD in Calgary, hemodialysis patients undergo at least 4 hours of dialysis three times weekly, aiming for a goal Kt/V of 1.3 or greater.26 Twenty percent of patients undergo treatment using high-flux dialyzers; no hemodialyzer reuse occurs. In 1999, there were 45 hemodialysis stations for in-center hemodialysis located in a tertiary hospital, 10 stations for satellite hemodialysis (located in the community), 3 stations for self-care hemodialysis, and 2 stations for home hemodialysis training. Patients on self-care hemodialysis operate independently of nurses. The training period for home and self-care hemodialysis (6 weeks while on dialysis) is similar. During 1999 to 2000, when this study was conducted, no formal hemodialysisaccess screening program was in place. Patients on peritoneal dialysis therapy undergo either continuous ambulatory peritoneal dialysis (CAPD) (usually
starting with four 2.5-L exchanges per day) or cyclic peritoneal dialysis (CCPD). All patients undergo a 1-week training period before the initiation of therapy. Treatment of peritoneal dialysis patients is individualized: the goal total Kt/V (renal and peritoneal clearance) is 2.1 or greater,27 and the majority of patients are treated with dextrose-based solutions. A minority of patients undergo a daily exchange with Nutrineal or Extraneal (Baxter Healthcare Corp, Deerfield, IL). The goal hemoglobin level locally for both hemodialysis and peritoneal dialysis patients is 110 to 120 g/L.
Data Collection Demographic information was assessed by means of a self-administered questionnaire. The presence of comorbid illness was assessed by a trained research nurse as of the enrollment date by direct patient interview and complete review of patients’ inpatient and outpatient records. Information was collected for the 19 variables that constitute the Charlson Comorbidity Index,28 which has been validated for use in patients with ESRD.29 Data for number of clinic visits, medication use, allocated nursing time for outpatient dialysis runs, and use of laboratory testing was determined for all participants from the SARP computerized database.30 Laboratory values considered in baseline characteristics were arithmetic means of hemoglobin, serum albumin, and Kt/V values measured during the 3-month period preceding enrollment.
Measuring Costs This study took the perspective of the health care purchaser and included only direct health care–related costs. Societal costs (ie, time or patient transport costs) were excluded. Although costs were measured in Canadian dollars, costs are reported here in 2000 US dollars (US $1 ⫽ CAN $1.45 ⫽ British £0.68). Costs were organized into four main categories: outpatient dialysis expenses, inpatient expenses, outpatient nondialysis expenses, and physician fees (see Appendix 1 for details). Outpatient dialysis comprises the largest portion of ESRD costs and includes equipment costs, staff, consumable items, and reverse-osmosis water (Appendix 1). To estimate inpatient expenses, information on all hospital admissions, including cost, was acquired from the Calgary Health Region (CHR) corporate database. Resource use was divided into seven categories: nursing, laboratory, diagnostic, surgical, surgical supplies, medications, and support staff (Appendix 1). Outpatient nondialysis expenses included clinic visits, emergency room, day surgery, laboratory tests, radiology, and medications (Appendix 1). Physician claims were acquired from Alberta Health and Wellness. Given that some costs considered may vary for different centers or may be uncertain, sensitivity analysis was used to determine whether changes in estimates of certain costs (physician claims, overhead rates charged to different modalities, nursing salary, hospitalization rates, and hospital costs) would significantly change the relative cost of different dialysis modalities. Because the goal of this study is to determine the cost of caring for patients treated with each of the different dialysis modalities, patients were analyzed according to the modality
THE COST OF DIALYSIS
with which they started the study; costs incurred after a change in modality or transplantation were censored.
Costs Related to Dialysis Access For hemodialysis patients, we assessed the cost of maintaining the vascular access for patients treated at the start of the study period with a permanent catheter, synthetic graft, or native AVF. Because we were interested in the cost of maintaining the dialysis access, patients who began the study with a temporary catheter (n ⫽ 5) were excluded from this analysis (all 166 patients, including these 5, were included in the dialysis modality cost analysis). Costs considered included any interventional radiology (ie, catheter placement, fistulograms, access angioplasty), day surgery for declotting and creation of new access, or hospital admission for access-related problems (defined as catheter sepsis or admission for an access-related surgical complication). Access-related physician costs are not included because we were unable to disaggregate physician claims based on service indication. In addition, we did not collect information on costs specifically related to outpatient use of intravenous antibiotics for access-related infection.
Statistical Analysis The primary study hypothesis is that the total cost of care for patients treated with the various dialytic modalities would be equivalent after controlling for other important variables. Descriptive statistics used were the arithmetic mean and 95% confidence intervals (CIs) for normally distributed samples and median with interquartile ranges for skewed variables. The Charlson index was calculated using the original formulae and weights.28 Two-sample two-sided t-tests or one-way analysis of variance (ANOVA) was used to test for differences in continuous variables between two or more groups, respectively. Chi-square was used to test for differences in discrete variables. The Kruskal-Wallis test was used to compare differences between highly skewed variables. Multiple linear regression was performed for the total cost of care, using the following predictor variables: age, sex, need for assisted living, work status, education, Charlson comorbidity score, and dialysis modality. Manual backward stepwise elimination of the least significant variable was performed based on a combination of clinical and statistical importance. Interaction and confounding were assessed for, and normal plots of residuals of regression models were examined to test for constant variance. We computed adjusted costs for average hypothetical patients treated with the various dialysis modalities (adjusted for significant predictor variables) using the least squares means method.
RESULTS
Patients Of 332 patients eligible for the study, 166 patients were enrolled; 88 patients were treated with in-center hemodialysis; 31 patients, satellite hemodialysis; 9 patients, home/self-care hemodialysis; and 38 patients, peritoneal dialysis (32
613
patients, CAPD; 6 patients, CCPD). One hundred ten patients were not enrolled because of inability to speak English or inability to contact them after repeated attempts, and 56 patients who were contacted did not wish to participate. Nonenrolled patients were similar in sex (55.6% men), but tended to be older (mean age, 63.8 years) and less likely to be on peritoneal dialysis therapy (16.4% of nonenrolled patients were on peritoneal dialysis therapy). Patients treated with in-center and satellite hemodialysis were older than those treated with self-care dialysis (home/self-care hemodialysis and peritoneal dialysis; Table 1). Patients treated with self-care dialysis were more likely to be working than those treated with in-center or satellite hemodialysis (P ⫽ 0.08, chi-square). Patients on home/self-care hemodialysis had less comorbid disease than patients treated with the other modalities (Table 1). For hemodialysis patients, vascular access was a temporary catheter in 3.9%, permanent catheter in 11.7%, synthetic graft in 28.1%, and native AVF in 56.3%. Costs We were able to assess costs for the full 12-month observation period for 124 of the 166 patients. Of the remainder, 16 patients died, 9 patients received a renal transplant, 4 patients switched dialysis modality (1 patient changed from in-center hemodialysis to peritoneal dialysis, and 3 patients switched from peritoneal dialysis to in-center hemodialysis), and 13 patients were enrolled after the start of the fixed costing period. Because we wanted to establish an accurate cost of treating a patient with a specific dialysis modality, for these patients, only a portion of the 12-month period was costed. Annual costs of treatment on each modality then were determined by direct extrapolation from this truncated costing period. Itemized health care costs for patients treated with each of the dialysis modalities are listed in Table 2. Overall annual costs of care for incenter, satellite and home/self-care hemodialysis and peritoneal dialysis were US $51,252 (95% CI, 47,680 to 54,824), $42,057 (95% CI, 39,523 to 44,592), $29,961 (95% CI, 21,252 to 38,670), and $26,959 (95% CI, 23,500 to 30,416), respectively (P ⬍ 0.001). After adjusting cost to account for the effect of the Charlson comorbidity
614
LEE ET AL Table 1.
Baseline Characteristics of Enrolled Patients Overall and for Patients Treated With Different Dialytic Modalities Hemodialysis
Variable Mean age* (y) Male sex* (%) White* (%) Working (%) Income (n ⫽ 130† respondents) (%) ⬍$10,000 10,000-30,000 ⬎30,000 Living arrangements (n ⫽ 157)† (%) Independent Require parttime home care Requires 24-h/long-term care Education (n ⫽ 159) (%) Less than grade 12 Grade 12/equivalent Postsecondary Median months on dialysis*‡ Mean Charlson comorbidity index* Low risk (ⱕ3) Medium risk (4-5) High risk (ⱖ6) Comorbid conditions (%) Coronary heart disease Congestive heart failure Peripheral vascular disease Previous stroke Lung disease Diabetes Mean baseline laboratory values Hemoglobin (g/dL) Albumin (g/dL) Kt/V
All Patients (n ⫽ 166)
In-Center (n ⫽ 88)
Satellite (n ⫽ 31)
Home/Self-Care (n ⫽ 9)
Peritoneal Dialysis (n ⫽ 38)
60.9 (58.5-63.3) 91 (54.8) 124 (74.7) 28 (16.9)
61.8 (58.4-65.3) 55.7 76.1 11.4
63.7 (58.4-69.1) 61.3 71.0 12.9
55.6 (43.3-67.9) 44.4 88.9 33.3
57.7 (52.6-62.9) 50.0 71.1 29.0
26 (20.0) 47 (36.2) 57 (43.9)
20.9 38.8 40.3
9.5 47.6 42.9
25.0 37.5 37.5
23.5 23.5 52.9
97 (61.8) 48 (30.6) 12 (7.6)
61.5 28.9 9.6
65.5 27.6 6.9
66.7 33.3 0.0
58.3 36.1 5.6
36 (22.6) 64 (40.3) 59 (37.1) 38.5 (31.9-45.0) 4.2 (3.9-4.5) 68 (41.0) 60 (36.1) 38 (22.9)
17.6 40.0 42.4 41.2 (31.9-50.5) 4.3 (3.9-4.7) 38.6 36.4 25.0
24.1 37.9 37.9 28.8 (23.6-34.1) 3.8 (3.2-4.5) 51.6 25.8 22.6
56 (33.7) 33 (19.9) 42 (25.30) 29 (17.47) 43 (25.90) 50 (30.12)
34.1 25.0 25.0 12.5 27.3 27.3
32.3 6.5 19.4 32.3 25.8 19.4
11.1 11.1 22.2 11.1 11.1 33.3
39.5 21.1 31.6 18.4 26.3 44.7
11.8 (11.6-12.0) 3.35 (3.29-3.42) 1.8 (1.7-1.9)
11.8 (11.6-12.1) 3.44 (3.36-3.51) 1.5 (1.5-1.6)
11.7 (11.3-12.0) 3.47 (3.36-3.57) 1.6 (1.5-1.6)
12.1 (11.3-12.9) 3.47 (3.26-3.67) 1.7 (1.5-1.9)
11.7 (11.1-12.4) 3.04 (2.88-3.19) 2.5 (2.3-2.7)
33.3 44.4 22.2 54.7 (11.4-120.8) 3.4 (2.5-4.4) 44.4 55.6 0.0
30.6 41.6 27.8 36.2 (22.2-50.1) 4.3 (3.7-5.0) 36.8 39.5 23.7
NOTE. Values expressed as number (percent) or mean (95% CI) unless noted otherwise. *P ⬍ 0.05 for comparison of dialytic modalities by one-way ANOVA. †N ⫽ 166 because of missing data. ‡Values expressed as median (interquartile range).
score (the only variable found to be independently associated with total costs, discussed later), adjusted costs for treating average hypothetical patients with in-center, satellite, and home/selfcare hemodialysis and peritoneal dialysis were US $50,928, $42,893, $31,679, and $26,540, respectively (P ⱕ 0.001). Table 3 lists a detailed breakdown of outpatient dialysis expenses. There was a greater requirement for nursing time for patients treated with in-center and satellite hemodialysis and greater use of supplies in patients treated with peritoneal dialysis. Outpatient dialysis costs were highest for in-center and satellite hemodialysis. For access-related costs (Table 4), patients who began the costing period with a functioning
AVF had significantly lower access-related costs compared with patients treated with a synthetic graft or permanent catheter (P ⬍ 0.001 by Kruskal-Wallis test). Sensitivity Analysis If physicians were remunerated equally for caring for patients treated with the different dialytic modalities, then the difference in cost of care for patients treated with in-center hemodialysis and self-care dialysis would be lower by $4,000/y. If nursing salaries were 30% higher, then the difference in cost between in-center hemodialysis and self-care dialysis would be higher by $2,500/y. If we considered 20% overhead for all dialytic modalities, then the differ-
THE COST OF DIALYSIS
Table 2.
Annual Health Care–Related Costs Per Patient by Modality Hemodialysis
Cost Category
Outpatient dialysis costs* ($) Inpatient costs*† ($) Outpatient nondialysis costs ($) Clinic visits* Emergency room* Day surgery* Laboratory costs* Diagnostic imaging* Erythropoietin*‡ Venofer* Medications*§ Total outpatient costs㛳 Physician billing* Total expenses㛳
In-Center (n ⫽ 88)
Satellite (n ⫽ 31)
Home/Self-Care (n ⫽ 9)
Peritoneal Dialysis (CAPD and CCPD) (n ⫽ 38)
26,692 (26,627, 26,282-26,982) 6,148 (0, 0-6,891)
24,113 (23,549, 23,326-23,947) 1,792 (0, 0-2,149)
14,282 (14,317, 13,018-15,472) 2,163 (0, 0-0)
12,494 (12,082, 11,406-13,210) 3,870 (0, 0-3,381)
43 (35, 25-61) 205 (70, 0-348) 236 (0, 0-0) 653 (524, 370-852) 1,703 (597, 99-2,168) 5,355 (4,037, 2,450-7,351) 381 (615, 0-615) 2,932 (2,612, 1,758-3,743) 11,418 (10,025-12,811) 6,995 (6,283, 6,283-7,412) 51,252 (47,680-54,824)
34 (32, 16-64) 157 (97, 0-252) 277 (0, 0-0) 332 (257, 190-391) 1,218 (519, 97-2,060) 4,310 (2,450, 0-7,351) 446 (615, 0-615) 2,616 (2,452, 1,859-3,386) 9,391 (7,540-11,242) 6,761 (6,283, 6,283-6,722) 42,057 (39,523-44,592)
41 (48, 16-48) 216 (0, 0-97) 197 (0, 0-422) 305 (288, 254-427) 1,701 (421, 200-2,626) 5,039 (4,901, 2,450-6,749) 404 (556, 308-615) 2,737 (2,482, 1,750-2,553) 10,639 (5,104-16,175) 2,877 (2,877, 2,877-2,877) 29,961 (21,252-38,670)
54 (48, 32-64) 352 (103, 0-341) 44 (0, 0-0) 439 (295, 208-616) 756 (97, 0-648) 4,054 (3,880, 1,125-7,351) 110 (0, 0-278) 2,941 (2,788, 1,460-3,618) 8,780 (7,278-10,283) 1,899 (1,550, 1,550-2,568) 26,959¶ (23,500-30,416)
NOTE. Values expressed as mean (median, 25th to 75th percentile) unless noted otherwise. *Reported as median with 25th to 75th percentile because distribution not normal. †Including nursing, medications, laboratory tests, diagnostic imaging, support staff, surgery, and supplies. ‡Locally, erythropoietin is administered only through the subcutaneous route. §Excluding erythropoietin and Venofer. 㛳Values expressed as mean (95% CI). ¶P ⬍ 0.001 comparing the four modalities using one-way ANOVA.
615
616
LEE ET AL Table 3.
Monthly Outpatient Dialysis Costs Per Patient by Modality Hemodialysis
Nursing*†‡ ($) Supplies ($) Machines ($) Water ($) Overhead㛳 ($) Total ($)
Peritoneal Dialysis
In-Center (n ⫽ 88)
Satellite (n ⫽ 31)
Home (n ⫽ 8)
Self Care (n ⫽ 1)
CAPD (n ⫽ 32)
CCPD (n ⫽ 6)
1,163 (1,157, 1,128-1,189) 421 96 31 513 2,224
1,027 (979, 961-1,013) 421 96 62 401 2,007
234 (227, 195-289) 421 146 243 209 1,253
320 (320, 320-320) 421 121 31 179 1,072
138 (114, 88-169) 862§ NA NA 200 1,200
102 (88, 59-156) 1,159§ NA NA 252 1,513¶
Abbreviation: NA, not applicable. *Including technician time. †Reported as median with 25th to 75th percentile because distribution was not normal. ‡Values expressed as mean (median, 25th to 75th percentile). §For CAPD, the average patient used four 2.5-L Dianeal exchanges per day, and for CCPD, the average patient cycled 10 L of Dianeal at night, with one 2-L Dianeal daytime exchange. 㛳Thirty percent for in-center, 25% for satellite and self-care, and 20% for home hemodialysis and peritoneal dialysis. ¶P ⬍ 0.001 for comparison of six dialytic modalities using one-way ANOVA.
ence in cost between in-center/satellite hemodialysis and self-care dialysis would be lower by $1,320/y. Using the detailed costing information listed in Tables 2 and 3, centers with different expected costs for any component of ESRD care could adjust the costs provided to determine a more accurate estimate of the local cost of each different dialytic modality. Predictors of Cost Using multiple linear regression, we found only Charlson comorbidity score and dialysis modality to be associated with health care costs (P ⬍ 0.001). Age, sex, need for assisted living, work status, and education status were not associated with annual costs (P ⬎ 0.10). For every increment in the Charlson comorbidity score by 1, annual costs increased by $2,234 (P ⬍ 0.001). Because one would not expect an equivalent increase in cost for each of the comorbidities that constitute the Charlson index, we also determined the effect of individual comorbid conditions on cost while controlling for the effect of other predictors. Patients with ESRD with diabetes had higher annual health care costs ($8,016; P ⬍ 0.001 compared with those without diabetes), as did those with a history of congestive heart failure ($6,873; P ⫽ 0.008 compared with those without a history of congestive heart failure).
DISCUSSION
Direct health care costs for patients with ESRD varied substantially among patients treated with different dialysis modalities. In-center hemodialysis was the most expensive renal replacement therapy, costing more than $50,000/patient-year. Self-care dialysis (ie, home/self-care hemodialysis and peritoneal dialysis) costs an average $20,000 less compared with in-center hemodialysis, in large part because of a lower requirement for nursing care for outpatient dialysis. Our results, based on patients treated with contemporary dialysis modalities in 1999 in North America, are consistent with previous work, which showed that home hemodialysis and peritoneal dialysis were the least expensive treatment options.6,15,19,24,31 This finding has now been reported in Europe,15,24 New Zealand,25 and Canada,6 although it was not as clear from these studies whether cost differences were caused by dialysis modality or differences in patient comorbidity between modalities. By controlling for the effect of patient comorbidity, our study shows that the lower cost of caring for patients treated with self-care dialysis was not caused by a difference in patient comorbidity. This result is consistent with a study by Hirth et al32 showing that the presence of additional comorbid disease did not significantly increase outpatient hemodialysis costs. It is apparent from our study that most of the cost savings associated with self-care dialysis
THE COST OF DIALYSIS
Table 4.
Annual Dialysis Access–Related Diagnostic Imaging, Surgical, and In-Patient Costs Excluding Physician Costs and Cost of Treating Outpatient Access-Associated Bacteremia
Access Type Hemodialysis㛳 Permanent catheter (n ⫽ 15) Synthetic graft (n ⫽ 36) Arteriovenous fistula (n ⫽ 72) Peritoneal dialysis Peritoneal dialysis catheter (n ⫽ 38)
Mean No./Patient/Y of Access-Related Interventional Radiology Visits*
Mean Access-Related Radiology Cost/ Patient/Y† ($)
Mean No./Patient/Y of Access-Related Surgical Visits‡
Mean AccessRelated Surgical Costs/Patient/ Y† ($)
Mean No./Patient/Y of Hospital Days for Access-Related Problems§
0.60 0.72 0.32
1,805 (1,010, 0-1,768) 2,017 (1,652, 0-3,120) 346 (0, 0-361)
0.20 1.08 0.10
85 (0, 0-0) 468 (0, 0-739) 58 (0, 0-0)
0.27 1.22 0
301 (0, 0-0) 860 (0, 0-0) 0
0.11
43 (0, 0-0)
0.44
179 (0, 0-0)
0.25
166 (0, 0-0)
Mean Cost/Patient/Y for Access-Related Hospital Admissions† ($)
Total Mean Costs/Patient/Y Costs for Access-Related Health Care† ($)
2,191 (1,479, 0-2,177) 3,345 (2,413, 407-4,638) 404 (0, 0-505)¶ 388 (0, 0-0)
*Includes vascular catheter placement, fistulograms, and access-related angioplasties. †Reported as median with 25th to 75th percentile because distribution was not normal. ‡Includes access declotting/revision and creation of new vascular access as a daycare patient. §Includes admission for catheter-associated bacteremia, other complications related to vascular access (postsurgical complications), and peritonitis. 㛳Patients who started the study with a temporary catheter (n ⫽ 5) who had active plans to create a permanent vascular access were excluded from this access-related cost analysis. ¶P ⬍ 0.001 comparing the three hemodialysis-access groups using the Kruskal-Wallis test.
617
618
was caused by a reduction in cost of outpatient dialysis (Table 2). These cost-savings are likely to be noted by other ESRD programs, irrespective of patient characteristics. A small proportion of the difference in cost that we noted between in-center or satellite hemodialysis and self-care dialysis was caused by a difference in the overhead considered for each modality (overhead is charged at a higher rate for in-center hemodialysis). In general, this is consistent with a lower use of such real expenses as laundry, housekeeping, portering, and electricity by home/self-care hemodialysis and peritoneal dialysis patients. Also, we did not consider any capital expenses (for building space) for the in-center, satellite, or self-care hemodialysis programs because it is difficult to estimate the cost of space within a hospital. However, using administrative data, we estimate the capital cost per in-center and satellite hemodialysis patient to be $723 and $296 per year, respectively. Perhaps more important than the monetary value, one should consider the opportunity cost of using space within a hospital for outpatient care of dialysis patients. If hospital space is a limited resource, as it is locally, then encouraging dialysis treatments outside of hospitals may be an important consideration, aside from monetary considerations. Although the discrepancy in costs is noteworthy, it also is important to consider health outcomes associated with treatment by the different dialysis modalities. As mentioned, previous work has shown no difference in survival9-12 or HRQOL13,15 between patients treated with hemodialysis or peritoneal dialysis. As such, it appears appropriate to put significant emphasis on comparing costs among the renal replacement therapies when determining which modality is most cost-efficient. Because self-care dialysis was least costly and also used less of the most scarce local resource (nursing time), we believe that for people who are awaiting or not eligible for a renal transplant, health care resources can be used most efficiently by encouraging treatment with either home/self-care hemodialysis or peritoneal dialysis. In the United States in 1999, a total of 243,320 patients were treated with dialysis.33 If the use of self-care dialysis were increased by 5% from the rate of 11.1% present in 1999 in the
LEE ET AL
United States,33 then more than US $243,000,000 could potentially be released annually to fund the treatment of new patients with ESRD or other effective treatments for existing patients with ESRD. Although the majority of patients with ESRD may still require treatment with in-center hemodialysis, we believe that the 5% increase in use of self-care dialysis proposed significantly underestimates the proportion of in-center hemodialysis patients who would be capable of performing self-care dialysis. At present, there may be economic incentives for dialysis providers in the United States and Canada to provide incenter rather than self-care dialysis2; we believe these incentives should be changed to encourage the appropriate use of self-care dialysis. Within hemodialysis, the cost of accessrelated care was significantly lower for those who began the study period with a functioning native AVF rather than a synthetic graft or permanent catheter. For patients eligible for a native AVF, our study suggests that in addition to extra clinical benefit in terms of longer access patency and lower risk for infection,34 there is the potential to realize substantial cost savings by encouraging placement of a native AVF rather than a synthetic graft. It will be important for future studies to consider the cost of access-related care from the time of access placement to account for patients who experience immediate access failure after AVF creation and subsequently require treatment with a permanent catheter or synthetic graft. In general, our results strongly support existing guidelines promoting the use of a native AVF as the first-line vascular access in hemodialysis patients with ESRD.35 It was interesting to note that the cost of maintaining hemodialysis patients with a synthetic graft was similar to that for permanent catheters. This may have been caused by the relatively higher rate of access intervention in those with synthetic grafts locally compared with results reported in programs in which access blood flow monitoring for synthetic grafts is routinely performed.36 Also, we did not collect costs specifically related to outpatient use of intravenous antibiotics for access-related infection, which would be expected to be highest for patients treated with a permanent catheter. Further research in this area will be important to clarify these issues.
THE COST OF DIALYSIS
This study had several limitations. We were able to enroll only 50% of eligible patients. We enrolled a limited number of peritoneal dialysis and home/self-care hemodialysis patients, although this is reflective of the local distribution of patients among the modalities. As such, it is possible that our results are affected by selection bias and enrolled patients were healthier than nonenrolled patients. If true, we may have underestimated costs, especially the cost of hospital admissions. For patients who were enrolled, by using the SARP database30 to collect data on allocated nursing time for outpatient dialysis runs, use of laboratory testing, and medication use, our estimate of costs of outpatient dialysis, erythropoietin, Venofer (American Regent Laboratories Inc, New York, NY), and outpatient medications were very accurate. It also is important to note that in Alberta, physician remuneration varies greatly between different modalities; it is highest for in-center and satellite dialysis. As such, physician cost per modality would need to be adjusted for regions in which nephrologists are paid on a capitation basis. For this analysis, we excluded such societal costs as time costs and those from work loss. Given that the proportion of patients working was greater for patients treated with self-care dialysis compared with in-center dialysis, if we had incorporated societal costs into our data, it is possible that the difference between self-care modalities and in-center hemodialysis would have been even greater. However, if patients who require significant assistance performing their dialysis were inappropriately treated with selfcare dialysis (which was not the case in our study), then inclusion of societal costs (including those of the caregiver) would make self-care dialysis appear less efficient. Regarding the generalizability of our results, it should be emphasized that patients on this study had been treated with their current modality for 6 months before enrollment, and all patients voluntarily chose their given modality. It is possible that the cost for each modality may differ if patients were coerced into treatment with a modality that they were unhappy with. Moreover, future research is required to determine start-up costs for each modality, which would better
619
reflect costs associated with switching dialysis modality, something most likely to occur during the first 6 months of dialysis treatment. Last, our study was performed in Canada; thus, its relevance to American centers must be addressed. It generally is accepted that hospitalization costs are higher in the United States compared with Canada and parts of Europe.2 However, given that the indication for hospitalization is generally unrelated to dialysis modality and tends to be related to patient comorbidity (ie, admissions for ischemic heart disease) or problems with vascular access (ie, synthetic graft declotting, infection of dialysis access), these higher hospitalization costs should not systematically bias the cost differences that we noted between the different modalities. (It is possible that the higher hospital costs may further exaggerate cost differences noted between maintaining a native AVF compared with a synthetic graft in the United States.) As we have shown, the relative difference in cost between different dialysis modalities generally relates to a difference in use of nursing time and supplies. Because nursing salaries and cost of medical care supplies are higher in the United States than Canada,37,38 the difference we noted in cost between in-center hemodialysis and self-care dialysis would be expected to be even higher in the United States. Because of the high cost of care and increasing prevalence of ESRD, it is important for ESRD programs to develop strategies for more efficient care. Although costing estimates in this study have most relevance to North American centers, we believe the general recommendation that can be made from our study (ie, that self-care dialysis should be promoted for eligible patients) is of relevance to other health systems. Potentially, resources released by patients performing selfcare dialysis (as opposed to in-center hemodialysis) could be used to enhance the care given to these same patients or other local patients with ESRD. APPENDIX 1
Costs were organized into four main categories: outpatient dialysis expenses, inpatient expenses, outpatient nondialysis expenses, and physician fees.
620
Outpatient Dialysis Expenses Outpatient dialysis expenses included equipment costs, staff, consumable items, and reverseosmosis water. The cost of the dialysis machine per run was determined as follows. Because 180 hemodialysis machines were required locally and provided 64,000 runs in 1999, a conservative estimate of the fixed cost per run was $7.39, assuming that: (1) an annuity factor of 7.3601 was used, (2) machines were amortized over 10 years with no resale value, and (3) maintenance costs for equipment (based on actual maintenance performed in 1999) were $2.12 per run. Staff costs included direct and indirect nursing costs, as well as the cost of such support staff as social workers, dietitians, and clerical staff. Direct nursing costs were estimated for each patient for every dialysis session based on a recorded workload measurement unit (WLM). Indirect nursing costs (ie, nursing administration) were divided equally among patients. Per standard CHR costing procedures, we used 30% overhead for in-center hemodialysis, 25% for satellite and self-care hemodialysis, and 20% for home hemodialysis and peritoneal dialysis. The cost of such consumable items as needles, blood catheters, and hemodialyzers for hemodialysis was based on average use (recorded in the monthly dialysis supplies records). Because reverse-osmosis water was purchased on a contract basis, we included the price of water required to reconstitute dialysate per patient. Inpatient Expenses Information on all hospital admissions (whether the admission was for reasons related or unrelated to kidney failure), including cost, was acquired from the CHR corporate database. Resource use was broken down into seven categories: nursing, laboratory, diagnostic, surgical, surgical supplies, medications, and support staff. In-patient costing was calculated using a provincially approved method in accordance with the Provincial and National Management Information Systems guidelines.39,40 Resource use (ie, number of medications, laboratory tests, or consumables used; nursing care hours allocated per patient based on a recorded WLM) was measured for all patients.30 The number of units used by each patient then was multiplied by the cost per unit, which is estimated annually in the CHR,
LEE ET AL
to give an estimate of direct costs. Direct costs then were combined with an estimate of indirect costs (defined as costs not directly associated with patient care, ie, cost of nursing managers, administration) for every in-patient encounter. The quality of data reporting of costs in Alberta has been highly ranked by the Canadian Institutes for Health Information.41 Because of the limited number of patients treated with CCPD, patients treated with CAPD and CCPD were pooled together to give a more stable estimate of inpatient costs. Inpatient costs were excluded from all other cost categories to avoid double counting. Outpatient Nondialysis Expenses Outpatient nondialysis expenses included clinic visits, emergency room, day surgery, laboratory tests, radiology, and medications. The cost of outpatient care was determined by examining actual patient resource consumption and allocated nursing contact for patients treated with in-center, satellite, home/self-care hemodialysis, and peritoneal dialysis. The SARP database30 records the number of clinic visits made by each patient; cost per visit was determined by dividing the annual clinic staff’s wages (including wages of clinic nurses, reception staff, and nursing managers) by total number of clinic visits annually for all patients and adding a 20% overhead charge. The number of outpatient visits to the emergency room, day surgery unit, and radiology was determined from the CHR corporate database. The cost for each visit (including indirect costs and 20% overhead) was estimated based on average cost for a representative sample of CHR patients who visited the emergency room (categorized by reason for assessment) or received treatment in day surgery or radiology (categorized by procedure). The frequency with which laboratory tests were performed was recorded for each patient by the SARP database.30 This frequency was multiplied by unit costs assigned to each laboratory test based on a defined WLM. Medications were divided into three categories: erythropoietin (Eprex; Ortho Biotech, Toronto, Ontario, Canada), intravenous iron (Venofer), and others. Frequency and dosage were extracted from the SARP database for each patient.30 For erythropoietin and intravenous iron,
THE COST OF DIALYSIS
we multiplied the dosage by unit costs (CAN $18.45 [US $12.71] per 1,000 units for Eprex and CAN $30 (US $20.67) per 100 mg for Venofer) to determine cost per patient. Locally, all patients administered erythropoietin do so through the subcutaneous route. All other medication costs were determined from the Alberta Pharmaceutical Price List, summed to give a total cost per patient. In Alberta, the cost of all medications for patients with ESRD is covered through a government-sponsored insurance plan (except for a $25/mon maximum copayment). Physician Fees To maintain patient confidentiality, physician claims were acquired from Alberta Health and Wellness in aggregate groups of 6 to 10 patients, according to dialysis modality and comorbidity. Physician claims included visits to general practitioner, nephrologist, radiologist, and other subspecialists. REFERENCES 1. US Renal Data System: Economic costs of ESRD. Am J Kidney Dis 36:S163-S176, 2000 (suppl 2) 2. Garella S: The costs of dialysis in the USA. Nephrol Dial Transplant 12:10-21, 1997 3. Mallick NP: The costs of renal services in Britain. Nephrol Dial Transplant 12:25-28, 1997 4. Tomson CR: Recent advances: Nephrology. BMJ 320: 98-101, 2000 5. US Renal Data System: Incidence and prevalence of ESRD. Am J Kidney Dis 30:S40-S53, 1997 (suppl 1) 6. Goeree R, Manalich J, Grootendorst P, Beecroft ML, Churchill DN: Cost analysis of dialysis treatments for endstage renal disease (ESRD). Clin Invest Med 18:455-464, 1995 7. Wolfe RA, Ashby VB, Milford EL, et al: Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med 341:1725-1730, 1999 8. Laupacis A, Keown P, Pus N, et al: A study of the quality of life and cost-utility of renal transplantation. Kidney Int 50:235-242, 1996 9. Bloembergen WE, Port FK, Mauger EA, Wolfe RA: A comparison of mortality between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 6:177183, 1995 10. Collins AJ, Hao W, Xia H, et al: Mortality risks of peritoneal dialysis and hemodialysis. Am J Kidney Dis 34:1065-1074, 1999 11. Fenton SS, Schaubel DE, Desmeules M, et al: Hemodialysis versus peritoneal dialysis: A comparison of adjusted mortality rates. Am J Kidney Dis 30:334-342, 1997 12. Foley RN, Parfrey PS, Harnett JD, et al: Mode of dialysis therapy and mortality in end-stage renal disease. J Am Soc Nephrol 9:267-276, 1998
621
13. Merkus MP, Jager KJ, Dekker FW, Boeschoten EW, Stevens P, Krediet RT: Quality of life in patients on chronic dialysis: Self-assessment 3 months after the start of treatment. The Necosad Study Group. Am J Kidney Dis 29:584592, 1997 14. Churchill DN, Torrance GW, Taylor DW, et al: Measurement of quality of life in end-stage renal disease: The time trade-off approach. Clin Invest Med 10:14-20, 1987 15. de Wit GA, Ramsteijn PG, de Charro FT: Economic evaluation of end stage renal disease treatment. Health Policy 44:215-232, 1998 16. Guernsey L: Self-care for the ESRD patient. Nephrol Nurse 4:8-12, 1982 17. Sexton Dobby A: Self-care hemodialysis training and anatomy: The patient’s perspective. CANNT J 2(4):17-18, 1992 18. Prichard SS: Treatment modality selection in 150 consecutive patients starting ESRD therapy. Perit Dial Int 16:69-72, 1996 19. Churchill DN, Lemon BC, Torrance GW: A costeffectiveness analysis of continuous ambulatory peritoneal dialysis and hospital hemodialysis. Med Decis Making 4:489500, 1984 20. Garner TI, Dardis R: Cost-effectiveness analysis of end-stage renal disease treatments. Med Care 25:25-34, 1987 21. Tousignant P, Guttmann RD, Hollomby DJ: Transplantation and home hemodialysis: Their cost-effectiveness. J Chronic Dis 38:589-601, 1985 22. Ludbrook A: A cost-effectiveness analysis of the treatment of chronic renal failure. Appl Econ 13:337-350, 1981 23. Thompson K: Cost comparison of in-center dialysis, home dialysis, and transplantation in the Nashville area. Nephrology Nurse 2(6):24-25, 1980 24. Weydevelt FC, Bacquaert-Dufour K, Benevent D, et al: A cost-effectiveness analysis of continuous ambulatory peritoneal dialysis vs self-care in-center hemodialysis in France. Dial Transplant 28:70-74, 1999 25. Croxson BE, Ashton T: A cost effectiveness analysis of the treatment of end stage renal failure. N Z Med J 103:171-174, 1990 26. Deziel C, Hirsch DJ, Hoult P, et al: Clinical practice guidelines for the delivery of hemodialysis. Canadian Society of Nephrology. J Am Soc Nephrol 10:S306-S310, 1999 (suppl 13) 27. Churchill D, Taylor DW, Keshaviah P: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7:198207, 1996 28. Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 40:373-383, 1987 29. Beddhu S, Bruns FJ, Saul M, Seddon P, Zeidel ML: A simple comorbidity scale predicts clinical outcomes and costs in dialysis patients. Am J Med 108:609-613, 2000 30. Manns BJ, Mortis G, Taub K, McLaughlin K, Donaldson C, Ghali WA: The Southern Alberta Renal Program Database: A prototype for patient management and research initiatives. Clin Invest Med 24:164-170, 2001
622
31. MacLeod A, Grant A, Donaldson C, et al: Effectiveness and efficiency of methods of dialysis therapy for end-stage renal disease: Systematic reviews. Health Technol Assess 2:1-166, 1998 32. Hirth RA, Held PJ, Orzol SM, Dor A: Practice patterns, case mix, Medicare payment policy, and dialysis facility costs. Health Serv Res 33:1567-1592, 1999 33. US Renal Data System: 2001 Atlas of ESRD in the United States, in (vol 2002), United States Renal Data System, 2001 34. Nassar GM, Ayus JC: Infectious complications of the hemodialysis access. Kidney Int 60:1-13, 2001 35. National Kidney Foundation: K/DOQI: Clinical Practice Guidelines for Vascular Access: Update 2000. Am J Kidney Dis 37:S137-S181, 2001 (suppl 1) 36. Schwab SJ, Oliver MJ, Suhocki P, McCann R: Hemodialysis arteriovenous access: Detection of stenosis and response to treatment by vascular access blood flow. Kidney Int 59:358-362, 2001
LEE ET AL
37. World Bank: World Development Indicators 2000. Washington, DC, World Bank, 2001 38. Feachem RG, Sekhri NK, White KL: Getting more for their dollar: A comparison of the NHS with California’s Kaiser Permanente. BMJ 324:135-141, 2002 39. Canadian Institute for Health Information: Guidelines for Management Information Systems in Canadian Health Service Organizations. Ottawa, Canada, Canadian Institute for Health Information, 1999 40. McKillop I: A Research Project to Examine the Costing Methodologies Recommended in the MIS Guidelines. Ottawa, Canada, Canadian Institutes for Health Information, 1995, p 155 41. McKillop I, Pink GH, Johnson LM: The Financial Management of Acute Care in Canada: A Review of Funding, Performance Monitoring and Reporting Practices. Ottawa, Canada, Canadian Institute for Health Information, 2001, p 83
C
ARE OF PATIENTS with end-stage renal disease (ESRD) is resource intense, but also important because renal replacement therapy is life saving. It has been estimated that 2% to 3% of a developed country’s health care budget is spent caring for patients with ESRD, although only 0.07% of the population has ESRD.1-4 Because of high costs and the increasing prevalence of ESRD,1,5 it is important for ESRD programs to develop strategies for more cost-efficient care. The largest area of cost for ESRD programs relates to the provision of outpatient renal replacement therapy.6 There are different methods of renal replacement therapy, including renal transplantation, hemodialysis, and peritoneal dialysis. When determining which modality to encourage, it is important to consider the effect of each modality on survival, health-related quality of life (HRQOL), and resource use. Renal transplantation is associated with superior clinical outcomes, HRQOL, and cost savings compared with dialysis.7,8 Unfortunately, many patients are not candidates for renal transplantation or must wait for a transplant on dialysis therapy.1 Regarding dialytic therapies, survival appears to be similar between peritoneal dialysis and hemodialysis,9-12 and studies suggest that HRQOL is similar for
patients on in-center hemodialysis and peritoneal dialysis therapy.13 Patients treated with home/selfcare hemodialysis may have a slightly greater HRQOL.14,15 Although some patients may have a medical or social preference for one modality, a large proportion of patients starting dialysis therapy could be treated successfully with any of the modalities.16-18 For this group of patients, who From the Departments of Economics, Community Health Sciences, and Medicine, University of Calgary; Alberta Centre of Health Services Utilization Research, Edmonton, Alberta; and the Department of Medicine, Anaesthesiology, Community Health and Epidemiology, University of Saskatchewan, Saskatchewan, Canada. Received February 28, 2002; accepted in revised form April 24, 2002. Work performed at Foothills Medical Centre, Calgary, Alberta, Canada. Supported in part by The Alberta Heritage Foundation for Medical Research and The Center for Advancement of Health, Calgary Regional Health Authority, Calgary, Alberta, Canada. Address reprint requests to Braden Manns, MD, Foothills Medical Centre, 1403 29th St NW, Calgary, Alberta, Canada, T2N 2T9. E-mail: [email protected] © 2002 by the National Kidney Foundation, Inc. 0272-6386/02/4003-0022$35.00/0 doi:10.1053/ajkd.2002.34924
American Journal of Kidney Diseases, Vol 40, No 3 (September), 2002: pp 611-622
611
612
LEE ET AL
may have limited knowledge about kidney failure and its treatment, it is important to educate patients about which modality would be optimal for them, keeping in mind the overall resources of the program and cost of each modality. Previous cost studies are largely outdated19-22; have used aggregate costs or prices23; have been performed outside of North America, where practice patterns may be different24,25; or have studied dialytic modalities (intermittent peritoneal dialysis) not in use today.6 The primary objective of our study is to provide ESRD programs with an updated and accurate itemized list of costs and resources required to treat patients with ESRD undergoing dialysis and contrast differences in resources required for various dialysis modalities. Our secondary objective is to estimate separately the cost of maintaining the dialysis access (arteriovenous fistula [AVF], synthetic graft, or permanent catheter). METHODS
Setting The Conjoint Medical and Research Ethics Board at the University of Calgary (Alberta, Canada) approved the study protocol. In July 1999, a total of 416 patients were treated locally with dialysis (266 patients, in-center hemodialysis; 55 patients, satellite hemodialysis; 79 patients, peritoneal dialysis; 13 patients, home hemodialysis; and 3 patients, self-care hemodialysis). Three hundred thirty-two of these patients had been on dialysis therapy longer than 6 months and were eligible for enrollment. Six months was chosen because (1) by that time, dialysis modality and a permanent vascular access generally have been established, and (2) the goal of this study is to determine the cost of ongoing dialysis, rather than costs associated with initiating dialysis therapy. Patients were approached for enrollment in a consecutive fashion from a randomly generated list of patients treated with each modality. Within the Southern Alberta Renal Program (SARP), which cares for all patients with ESRD in Calgary, hemodialysis patients undergo at least 4 hours of dialysis three times weekly, aiming for a goal Kt/V of 1.3 or greater.26 Twenty percent of patients undergo treatment using high-flux dialyzers; no hemodialyzer reuse occurs. In 1999, there were 45 hemodialysis stations for in-center hemodialysis located in a tertiary hospital, 10 stations for satellite hemodialysis (located in the community), 3 stations for self-care hemodialysis, and 2 stations for home hemodialysis training. Patients on self-care hemodialysis operate independently of nurses. The training period for home and self-care hemodialysis (6 weeks while on dialysis) is similar. During 1999 to 2000, when this study was conducted, no formal hemodialysisaccess screening program was in place. Patients on peritoneal dialysis therapy undergo either continuous ambulatory peritoneal dialysis (CAPD) (usually
starting with four 2.5-L exchanges per day) or cyclic peritoneal dialysis (CCPD). All patients undergo a 1-week training period before the initiation of therapy. Treatment of peritoneal dialysis patients is individualized: the goal total Kt/V (renal and peritoneal clearance) is 2.1 or greater,27 and the majority of patients are treated with dextrose-based solutions. A minority of patients undergo a daily exchange with Nutrineal or Extraneal (Baxter Healthcare Corp, Deerfield, IL). The goal hemoglobin level locally for both hemodialysis and peritoneal dialysis patients is 110 to 120 g/L.
Data Collection Demographic information was assessed by means of a self-administered questionnaire. The presence of comorbid illness was assessed by a trained research nurse as of the enrollment date by direct patient interview and complete review of patients’ inpatient and outpatient records. Information was collected for the 19 variables that constitute the Charlson Comorbidity Index,28 which has been validated for use in patients with ESRD.29 Data for number of clinic visits, medication use, allocated nursing time for outpatient dialysis runs, and use of laboratory testing was determined for all participants from the SARP computerized database.30 Laboratory values considered in baseline characteristics were arithmetic means of hemoglobin, serum albumin, and Kt/V values measured during the 3-month period preceding enrollment.
Measuring Costs This study took the perspective of the health care purchaser and included only direct health care–related costs. Societal costs (ie, time or patient transport costs) were excluded. Although costs were measured in Canadian dollars, costs are reported here in 2000 US dollars (US $1 ⫽ CAN $1.45 ⫽ British £0.68). Costs were organized into four main categories: outpatient dialysis expenses, inpatient expenses, outpatient nondialysis expenses, and physician fees (see Appendix 1 for details). Outpatient dialysis comprises the largest portion of ESRD costs and includes equipment costs, staff, consumable items, and reverse-osmosis water (Appendix 1). To estimate inpatient expenses, information on all hospital admissions, including cost, was acquired from the Calgary Health Region (CHR) corporate database. Resource use was divided into seven categories: nursing, laboratory, diagnostic, surgical, surgical supplies, medications, and support staff (Appendix 1). Outpatient nondialysis expenses included clinic visits, emergency room, day surgery, laboratory tests, radiology, and medications (Appendix 1). Physician claims were acquired from Alberta Health and Wellness. Given that some costs considered may vary for different centers or may be uncertain, sensitivity analysis was used to determine whether changes in estimates of certain costs (physician claims, overhead rates charged to different modalities, nursing salary, hospitalization rates, and hospital costs) would significantly change the relative cost of different dialysis modalities. Because the goal of this study is to determine the cost of caring for patients treated with each of the different dialysis modalities, patients were analyzed according to the modality
THE COST OF DIALYSIS
with which they started the study; costs incurred after a change in modality or transplantation were censored.
Costs Related to Dialysis Access For hemodialysis patients, we assessed the cost of maintaining the vascular access for patients treated at the start of the study period with a permanent catheter, synthetic graft, or native AVF. Because we were interested in the cost of maintaining the dialysis access, patients who began the study with a temporary catheter (n ⫽ 5) were excluded from this analysis (all 166 patients, including these 5, were included in the dialysis modality cost analysis). Costs considered included any interventional radiology (ie, catheter placement, fistulograms, access angioplasty), day surgery for declotting and creation of new access, or hospital admission for access-related problems (defined as catheter sepsis or admission for an access-related surgical complication). Access-related physician costs are not included because we were unable to disaggregate physician claims based on service indication. In addition, we did not collect information on costs specifically related to outpatient use of intravenous antibiotics for access-related infection.
Statistical Analysis The primary study hypothesis is that the total cost of care for patients treated with the various dialytic modalities would be equivalent after controlling for other important variables. Descriptive statistics used were the arithmetic mean and 95% confidence intervals (CIs) for normally distributed samples and median with interquartile ranges for skewed variables. The Charlson index was calculated using the original formulae and weights.28 Two-sample two-sided t-tests or one-way analysis of variance (ANOVA) was used to test for differences in continuous variables between two or more groups, respectively. Chi-square was used to test for differences in discrete variables. The Kruskal-Wallis test was used to compare differences between highly skewed variables. Multiple linear regression was performed for the total cost of care, using the following predictor variables: age, sex, need for assisted living, work status, education, Charlson comorbidity score, and dialysis modality. Manual backward stepwise elimination of the least significant variable was performed based on a combination of clinical and statistical importance. Interaction and confounding were assessed for, and normal plots of residuals of regression models were examined to test for constant variance. We computed adjusted costs for average hypothetical patients treated with the various dialysis modalities (adjusted for significant predictor variables) using the least squares means method.
RESULTS
Patients Of 332 patients eligible for the study, 166 patients were enrolled; 88 patients were treated with in-center hemodialysis; 31 patients, satellite hemodialysis; 9 patients, home/self-care hemodialysis; and 38 patients, peritoneal dialysis (32
613
patients, CAPD; 6 patients, CCPD). One hundred ten patients were not enrolled because of inability to speak English or inability to contact them after repeated attempts, and 56 patients who were contacted did not wish to participate. Nonenrolled patients were similar in sex (55.6% men), but tended to be older (mean age, 63.8 years) and less likely to be on peritoneal dialysis therapy (16.4% of nonenrolled patients were on peritoneal dialysis therapy). Patients treated with in-center and satellite hemodialysis were older than those treated with self-care dialysis (home/self-care hemodialysis and peritoneal dialysis; Table 1). Patients treated with self-care dialysis were more likely to be working than those treated with in-center or satellite hemodialysis (P ⫽ 0.08, chi-square). Patients on home/self-care hemodialysis had less comorbid disease than patients treated with the other modalities (Table 1). For hemodialysis patients, vascular access was a temporary catheter in 3.9%, permanent catheter in 11.7%, synthetic graft in 28.1%, and native AVF in 56.3%. Costs We were able to assess costs for the full 12-month observation period for 124 of the 166 patients. Of the remainder, 16 patients died, 9 patients received a renal transplant, 4 patients switched dialysis modality (1 patient changed from in-center hemodialysis to peritoneal dialysis, and 3 patients switched from peritoneal dialysis to in-center hemodialysis), and 13 patients were enrolled after the start of the fixed costing period. Because we wanted to establish an accurate cost of treating a patient with a specific dialysis modality, for these patients, only a portion of the 12-month period was costed. Annual costs of treatment on each modality then were determined by direct extrapolation from this truncated costing period. Itemized health care costs for patients treated with each of the dialysis modalities are listed in Table 2. Overall annual costs of care for incenter, satellite and home/self-care hemodialysis and peritoneal dialysis were US $51,252 (95% CI, 47,680 to 54,824), $42,057 (95% CI, 39,523 to 44,592), $29,961 (95% CI, 21,252 to 38,670), and $26,959 (95% CI, 23,500 to 30,416), respectively (P ⬍ 0.001). After adjusting cost to account for the effect of the Charlson comorbidity
614
LEE ET AL Table 1.
Baseline Characteristics of Enrolled Patients Overall and for Patients Treated With Different Dialytic Modalities Hemodialysis
Variable Mean age* (y) Male sex* (%) White* (%) Working (%) Income (n ⫽ 130† respondents) (%) ⬍$10,000 10,000-30,000 ⬎30,000 Living arrangements (n ⫽ 157)† (%) Independent Require parttime home care Requires 24-h/long-term care Education (n ⫽ 159) (%) Less than grade 12 Grade 12/equivalent Postsecondary Median months on dialysis*‡ Mean Charlson comorbidity index* Low risk (ⱕ3) Medium risk (4-5) High risk (ⱖ6) Comorbid conditions (%) Coronary heart disease Congestive heart failure Peripheral vascular disease Previous stroke Lung disease Diabetes Mean baseline laboratory values Hemoglobin (g/dL) Albumin (g/dL) Kt/V
All Patients (n ⫽ 166)
In-Center (n ⫽ 88)
Satellite (n ⫽ 31)
Home/Self-Care (n ⫽ 9)
Peritoneal Dialysis (n ⫽ 38)
60.9 (58.5-63.3) 91 (54.8) 124 (74.7) 28 (16.9)
61.8 (58.4-65.3) 55.7 76.1 11.4
63.7 (58.4-69.1) 61.3 71.0 12.9
55.6 (43.3-67.9) 44.4 88.9 33.3
57.7 (52.6-62.9) 50.0 71.1 29.0
26 (20.0) 47 (36.2) 57 (43.9)
20.9 38.8 40.3
9.5 47.6 42.9
25.0 37.5 37.5
23.5 23.5 52.9
97 (61.8) 48 (30.6) 12 (7.6)
61.5 28.9 9.6
65.5 27.6 6.9
66.7 33.3 0.0
58.3 36.1 5.6
36 (22.6) 64 (40.3) 59 (37.1) 38.5 (31.9-45.0) 4.2 (3.9-4.5) 68 (41.0) 60 (36.1) 38 (22.9)
17.6 40.0 42.4 41.2 (31.9-50.5) 4.3 (3.9-4.7) 38.6 36.4 25.0
24.1 37.9 37.9 28.8 (23.6-34.1) 3.8 (3.2-4.5) 51.6 25.8 22.6
56 (33.7) 33 (19.9) 42 (25.30) 29 (17.47) 43 (25.90) 50 (30.12)
34.1 25.0 25.0 12.5 27.3 27.3
32.3 6.5 19.4 32.3 25.8 19.4
11.1 11.1 22.2 11.1 11.1 33.3
39.5 21.1 31.6 18.4 26.3 44.7
11.8 (11.6-12.0) 3.35 (3.29-3.42) 1.8 (1.7-1.9)
11.8 (11.6-12.1) 3.44 (3.36-3.51) 1.5 (1.5-1.6)
11.7 (11.3-12.0) 3.47 (3.36-3.57) 1.6 (1.5-1.6)
12.1 (11.3-12.9) 3.47 (3.26-3.67) 1.7 (1.5-1.9)
11.7 (11.1-12.4) 3.04 (2.88-3.19) 2.5 (2.3-2.7)
33.3 44.4 22.2 54.7 (11.4-120.8) 3.4 (2.5-4.4) 44.4 55.6 0.0
30.6 41.6 27.8 36.2 (22.2-50.1) 4.3 (3.7-5.0) 36.8 39.5 23.7
NOTE. Values expressed as number (percent) or mean (95% CI) unless noted otherwise. *P ⬍ 0.05 for comparison of dialytic modalities by one-way ANOVA. †N ⫽ 166 because of missing data. ‡Values expressed as median (interquartile range).
score (the only variable found to be independently associated with total costs, discussed later), adjusted costs for treating average hypothetical patients with in-center, satellite, and home/selfcare hemodialysis and peritoneal dialysis were US $50,928, $42,893, $31,679, and $26,540, respectively (P ⱕ 0.001). Table 3 lists a detailed breakdown of outpatient dialysis expenses. There was a greater requirement for nursing time for patients treated with in-center and satellite hemodialysis and greater use of supplies in patients treated with peritoneal dialysis. Outpatient dialysis costs were highest for in-center and satellite hemodialysis. For access-related costs (Table 4), patients who began the costing period with a functioning
AVF had significantly lower access-related costs compared with patients treated with a synthetic graft or permanent catheter (P ⬍ 0.001 by Kruskal-Wallis test). Sensitivity Analysis If physicians were remunerated equally for caring for patients treated with the different dialytic modalities, then the difference in cost of care for patients treated with in-center hemodialysis and self-care dialysis would be lower by $4,000/y. If nursing salaries were 30% higher, then the difference in cost between in-center hemodialysis and self-care dialysis would be higher by $2,500/y. If we considered 20% overhead for all dialytic modalities, then the differ-
THE COST OF DIALYSIS
Table 2.
Annual Health Care–Related Costs Per Patient by Modality Hemodialysis
Cost Category
Outpatient dialysis costs* ($) Inpatient costs*† ($) Outpatient nondialysis costs ($) Clinic visits* Emergency room* Day surgery* Laboratory costs* Diagnostic imaging* Erythropoietin*‡ Venofer* Medications*§ Total outpatient costs㛳 Physician billing* Total expenses㛳
In-Center (n ⫽ 88)
Satellite (n ⫽ 31)
Home/Self-Care (n ⫽ 9)
Peritoneal Dialysis (CAPD and CCPD) (n ⫽ 38)
26,692 (26,627, 26,282-26,982) 6,148 (0, 0-6,891)
24,113 (23,549, 23,326-23,947) 1,792 (0, 0-2,149)
14,282 (14,317, 13,018-15,472) 2,163 (0, 0-0)
12,494 (12,082, 11,406-13,210) 3,870 (0, 0-3,381)
43 (35, 25-61) 205 (70, 0-348) 236 (0, 0-0) 653 (524, 370-852) 1,703 (597, 99-2,168) 5,355 (4,037, 2,450-7,351) 381 (615, 0-615) 2,932 (2,612, 1,758-3,743) 11,418 (10,025-12,811) 6,995 (6,283, 6,283-7,412) 51,252 (47,680-54,824)
34 (32, 16-64) 157 (97, 0-252) 277 (0, 0-0) 332 (257, 190-391) 1,218 (519, 97-2,060) 4,310 (2,450, 0-7,351) 446 (615, 0-615) 2,616 (2,452, 1,859-3,386) 9,391 (7,540-11,242) 6,761 (6,283, 6,283-6,722) 42,057 (39,523-44,592)
41 (48, 16-48) 216 (0, 0-97) 197 (0, 0-422) 305 (288, 254-427) 1,701 (421, 200-2,626) 5,039 (4,901, 2,450-6,749) 404 (556, 308-615) 2,737 (2,482, 1,750-2,553) 10,639 (5,104-16,175) 2,877 (2,877, 2,877-2,877) 29,961 (21,252-38,670)
54 (48, 32-64) 352 (103, 0-341) 44 (0, 0-0) 439 (295, 208-616) 756 (97, 0-648) 4,054 (3,880, 1,125-7,351) 110 (0, 0-278) 2,941 (2,788, 1,460-3,618) 8,780 (7,278-10,283) 1,899 (1,550, 1,550-2,568) 26,959¶ (23,500-30,416)
NOTE. Values expressed as mean (median, 25th to 75th percentile) unless noted otherwise. *Reported as median with 25th to 75th percentile because distribution not normal. †Including nursing, medications, laboratory tests, diagnostic imaging, support staff, surgery, and supplies. ‡Locally, erythropoietin is administered only through the subcutaneous route. §Excluding erythropoietin and Venofer. 㛳Values expressed as mean (95% CI). ¶P ⬍ 0.001 comparing the four modalities using one-way ANOVA.
615
616
LEE ET AL Table 3.
Monthly Outpatient Dialysis Costs Per Patient by Modality Hemodialysis
Nursing*†‡ ($) Supplies ($) Machines ($) Water ($) Overhead㛳 ($) Total ($)
Peritoneal Dialysis
In-Center (n ⫽ 88)
Satellite (n ⫽ 31)
Home (n ⫽ 8)
Self Care (n ⫽ 1)
CAPD (n ⫽ 32)
CCPD (n ⫽ 6)
1,163 (1,157, 1,128-1,189) 421 96 31 513 2,224
1,027 (979, 961-1,013) 421 96 62 401 2,007
234 (227, 195-289) 421 146 243 209 1,253
320 (320, 320-320) 421 121 31 179 1,072
138 (114, 88-169) 862§ NA NA 200 1,200
102 (88, 59-156) 1,159§ NA NA 252 1,513¶
Abbreviation: NA, not applicable. *Including technician time. †Reported as median with 25th to 75th percentile because distribution was not normal. ‡Values expressed as mean (median, 25th to 75th percentile). §For CAPD, the average patient used four 2.5-L Dianeal exchanges per day, and for CCPD, the average patient cycled 10 L of Dianeal at night, with one 2-L Dianeal daytime exchange. 㛳Thirty percent for in-center, 25% for satellite and self-care, and 20% for home hemodialysis and peritoneal dialysis. ¶P ⬍ 0.001 for comparison of six dialytic modalities using one-way ANOVA.
ence in cost between in-center/satellite hemodialysis and self-care dialysis would be lower by $1,320/y. Using the detailed costing information listed in Tables 2 and 3, centers with different expected costs for any component of ESRD care could adjust the costs provided to determine a more accurate estimate of the local cost of each different dialytic modality. Predictors of Cost Using multiple linear regression, we found only Charlson comorbidity score and dialysis modality to be associated with health care costs (P ⬍ 0.001). Age, sex, need for assisted living, work status, and education status were not associated with annual costs (P ⬎ 0.10). For every increment in the Charlson comorbidity score by 1, annual costs increased by $2,234 (P ⬍ 0.001). Because one would not expect an equivalent increase in cost for each of the comorbidities that constitute the Charlson index, we also determined the effect of individual comorbid conditions on cost while controlling for the effect of other predictors. Patients with ESRD with diabetes had higher annual health care costs ($8,016; P ⬍ 0.001 compared with those without diabetes), as did those with a history of congestive heart failure ($6,873; P ⫽ 0.008 compared with those without a history of congestive heart failure).
DISCUSSION
Direct health care costs for patients with ESRD varied substantially among patients treated with different dialysis modalities. In-center hemodialysis was the most expensive renal replacement therapy, costing more than $50,000/patient-year. Self-care dialysis (ie, home/self-care hemodialysis and peritoneal dialysis) costs an average $20,000 less compared with in-center hemodialysis, in large part because of a lower requirement for nursing care for outpatient dialysis. Our results, based on patients treated with contemporary dialysis modalities in 1999 in North America, are consistent with previous work, which showed that home hemodialysis and peritoneal dialysis were the least expensive treatment options.6,15,19,24,31 This finding has now been reported in Europe,15,24 New Zealand,25 and Canada,6 although it was not as clear from these studies whether cost differences were caused by dialysis modality or differences in patient comorbidity between modalities. By controlling for the effect of patient comorbidity, our study shows that the lower cost of caring for patients treated with self-care dialysis was not caused by a difference in patient comorbidity. This result is consistent with a study by Hirth et al32 showing that the presence of additional comorbid disease did not significantly increase outpatient hemodialysis costs. It is apparent from our study that most of the cost savings associated with self-care dialysis
THE COST OF DIALYSIS
Table 4.
Annual Dialysis Access–Related Diagnostic Imaging, Surgical, and In-Patient Costs Excluding Physician Costs and Cost of Treating Outpatient Access-Associated Bacteremia
Access Type Hemodialysis㛳 Permanent catheter (n ⫽ 15) Synthetic graft (n ⫽ 36) Arteriovenous fistula (n ⫽ 72) Peritoneal dialysis Peritoneal dialysis catheter (n ⫽ 38)
Mean No./Patient/Y of Access-Related Interventional Radiology Visits*
Mean Access-Related Radiology Cost/ Patient/Y† ($)
Mean No./Patient/Y of Access-Related Surgical Visits‡
Mean AccessRelated Surgical Costs/Patient/ Y† ($)
Mean No./Patient/Y of Hospital Days for Access-Related Problems§
0.60 0.72 0.32
1,805 (1,010, 0-1,768) 2,017 (1,652, 0-3,120) 346 (0, 0-361)
0.20 1.08 0.10
85 (0, 0-0) 468 (0, 0-739) 58 (0, 0-0)
0.27 1.22 0
301 (0, 0-0) 860 (0, 0-0) 0
0.11
43 (0, 0-0)
0.44
179 (0, 0-0)
0.25
166 (0, 0-0)
Mean Cost/Patient/Y for Access-Related Hospital Admissions† ($)
Total Mean Costs/Patient/Y Costs for Access-Related Health Care† ($)
2,191 (1,479, 0-2,177) 3,345 (2,413, 407-4,638) 404 (0, 0-505)¶ 388 (0, 0-0)
*Includes vascular catheter placement, fistulograms, and access-related angioplasties. †Reported as median with 25th to 75th percentile because distribution was not normal. ‡Includes access declotting/revision and creation of new vascular access as a daycare patient. §Includes admission for catheter-associated bacteremia, other complications related to vascular access (postsurgical complications), and peritonitis. 㛳Patients who started the study with a temporary catheter (n ⫽ 5) who had active plans to create a permanent vascular access were excluded from this access-related cost analysis. ¶P ⬍ 0.001 comparing the three hemodialysis-access groups using the Kruskal-Wallis test.
617
618
was caused by a reduction in cost of outpatient dialysis (Table 2). These cost-savings are likely to be noted by other ESRD programs, irrespective of patient characteristics. A small proportion of the difference in cost that we noted between in-center or satellite hemodialysis and self-care dialysis was caused by a difference in the overhead considered for each modality (overhead is charged at a higher rate for in-center hemodialysis). In general, this is consistent with a lower use of such real expenses as laundry, housekeeping, portering, and electricity by home/self-care hemodialysis and peritoneal dialysis patients. Also, we did not consider any capital expenses (for building space) for the in-center, satellite, or self-care hemodialysis programs because it is difficult to estimate the cost of space within a hospital. However, using administrative data, we estimate the capital cost per in-center and satellite hemodialysis patient to be $723 and $296 per year, respectively. Perhaps more important than the monetary value, one should consider the opportunity cost of using space within a hospital for outpatient care of dialysis patients. If hospital space is a limited resource, as it is locally, then encouraging dialysis treatments outside of hospitals may be an important consideration, aside from monetary considerations. Although the discrepancy in costs is noteworthy, it also is important to consider health outcomes associated with treatment by the different dialysis modalities. As mentioned, previous work has shown no difference in survival9-12 or HRQOL13,15 between patients treated with hemodialysis or peritoneal dialysis. As such, it appears appropriate to put significant emphasis on comparing costs among the renal replacement therapies when determining which modality is most cost-efficient. Because self-care dialysis was least costly and also used less of the most scarce local resource (nursing time), we believe that for people who are awaiting or not eligible for a renal transplant, health care resources can be used most efficiently by encouraging treatment with either home/self-care hemodialysis or peritoneal dialysis. In the United States in 1999, a total of 243,320 patients were treated with dialysis.33 If the use of self-care dialysis were increased by 5% from the rate of 11.1% present in 1999 in the
LEE ET AL
United States,33 then more than US $243,000,000 could potentially be released annually to fund the treatment of new patients with ESRD or other effective treatments for existing patients with ESRD. Although the majority of patients with ESRD may still require treatment with in-center hemodialysis, we believe that the 5% increase in use of self-care dialysis proposed significantly underestimates the proportion of in-center hemodialysis patients who would be capable of performing self-care dialysis. At present, there may be economic incentives for dialysis providers in the United States and Canada to provide incenter rather than self-care dialysis2; we believe these incentives should be changed to encourage the appropriate use of self-care dialysis. Within hemodialysis, the cost of accessrelated care was significantly lower for those who began the study period with a functioning native AVF rather than a synthetic graft or permanent catheter. For patients eligible for a native AVF, our study suggests that in addition to extra clinical benefit in terms of longer access patency and lower risk for infection,34 there is the potential to realize substantial cost savings by encouraging placement of a native AVF rather than a synthetic graft. It will be important for future studies to consider the cost of access-related care from the time of access placement to account for patients who experience immediate access failure after AVF creation and subsequently require treatment with a permanent catheter or synthetic graft. In general, our results strongly support existing guidelines promoting the use of a native AVF as the first-line vascular access in hemodialysis patients with ESRD.35 It was interesting to note that the cost of maintaining hemodialysis patients with a synthetic graft was similar to that for permanent catheters. This may have been caused by the relatively higher rate of access intervention in those with synthetic grafts locally compared with results reported in programs in which access blood flow monitoring for synthetic grafts is routinely performed.36 Also, we did not collect costs specifically related to outpatient use of intravenous antibiotics for access-related infection, which would be expected to be highest for patients treated with a permanent catheter. Further research in this area will be important to clarify these issues.
THE COST OF DIALYSIS
This study had several limitations. We were able to enroll only 50% of eligible patients. We enrolled a limited number of peritoneal dialysis and home/self-care hemodialysis patients, although this is reflective of the local distribution of patients among the modalities. As such, it is possible that our results are affected by selection bias and enrolled patients were healthier than nonenrolled patients. If true, we may have underestimated costs, especially the cost of hospital admissions. For patients who were enrolled, by using the SARP database30 to collect data on allocated nursing time for outpatient dialysis runs, use of laboratory testing, and medication use, our estimate of costs of outpatient dialysis, erythropoietin, Venofer (American Regent Laboratories Inc, New York, NY), and outpatient medications were very accurate. It also is important to note that in Alberta, physician remuneration varies greatly between different modalities; it is highest for in-center and satellite dialysis. As such, physician cost per modality would need to be adjusted for regions in which nephrologists are paid on a capitation basis. For this analysis, we excluded such societal costs as time costs and those from work loss. Given that the proportion of patients working was greater for patients treated with self-care dialysis compared with in-center dialysis, if we had incorporated societal costs into our data, it is possible that the difference between self-care modalities and in-center hemodialysis would have been even greater. However, if patients who require significant assistance performing their dialysis were inappropriately treated with selfcare dialysis (which was not the case in our study), then inclusion of societal costs (including those of the caregiver) would make self-care dialysis appear less efficient. Regarding the generalizability of our results, it should be emphasized that patients on this study had been treated with their current modality for 6 months before enrollment, and all patients voluntarily chose their given modality. It is possible that the cost for each modality may differ if patients were coerced into treatment with a modality that they were unhappy with. Moreover, future research is required to determine start-up costs for each modality, which would better
619
reflect costs associated with switching dialysis modality, something most likely to occur during the first 6 months of dialysis treatment. Last, our study was performed in Canada; thus, its relevance to American centers must be addressed. It generally is accepted that hospitalization costs are higher in the United States compared with Canada and parts of Europe.2 However, given that the indication for hospitalization is generally unrelated to dialysis modality and tends to be related to patient comorbidity (ie, admissions for ischemic heart disease) or problems with vascular access (ie, synthetic graft declotting, infection of dialysis access), these higher hospitalization costs should not systematically bias the cost differences that we noted between the different modalities. (It is possible that the higher hospital costs may further exaggerate cost differences noted between maintaining a native AVF compared with a synthetic graft in the United States.) As we have shown, the relative difference in cost between different dialysis modalities generally relates to a difference in use of nursing time and supplies. Because nursing salaries and cost of medical care supplies are higher in the United States than Canada,37,38 the difference we noted in cost between in-center hemodialysis and self-care dialysis would be expected to be even higher in the United States. Because of the high cost of care and increasing prevalence of ESRD, it is important for ESRD programs to develop strategies for more efficient care. Although costing estimates in this study have most relevance to North American centers, we believe the general recommendation that can be made from our study (ie, that self-care dialysis should be promoted for eligible patients) is of relevance to other health systems. Potentially, resources released by patients performing selfcare dialysis (as opposed to in-center hemodialysis) could be used to enhance the care given to these same patients or other local patients with ESRD. APPENDIX 1
Costs were organized into four main categories: outpatient dialysis expenses, inpatient expenses, outpatient nondialysis expenses, and physician fees.
620
Outpatient Dialysis Expenses Outpatient dialysis expenses included equipment costs, staff, consumable items, and reverseosmosis water. The cost of the dialysis machine per run was determined as follows. Because 180 hemodialysis machines were required locally and provided 64,000 runs in 1999, a conservative estimate of the fixed cost per run was $7.39, assuming that: (1) an annuity factor of 7.3601 was used, (2) machines were amortized over 10 years with no resale value, and (3) maintenance costs for equipment (based on actual maintenance performed in 1999) were $2.12 per run. Staff costs included direct and indirect nursing costs, as well as the cost of such support staff as social workers, dietitians, and clerical staff. Direct nursing costs were estimated for each patient for every dialysis session based on a recorded workload measurement unit (WLM). Indirect nursing costs (ie, nursing administration) were divided equally among patients. Per standard CHR costing procedures, we used 30% overhead for in-center hemodialysis, 25% for satellite and self-care hemodialysis, and 20% for home hemodialysis and peritoneal dialysis. The cost of such consumable items as needles, blood catheters, and hemodialyzers for hemodialysis was based on average use (recorded in the monthly dialysis supplies records). Because reverse-osmosis water was purchased on a contract basis, we included the price of water required to reconstitute dialysate per patient. Inpatient Expenses Information on all hospital admissions (whether the admission was for reasons related or unrelated to kidney failure), including cost, was acquired from the CHR corporate database. Resource use was broken down into seven categories: nursing, laboratory, diagnostic, surgical, surgical supplies, medications, and support staff. In-patient costing was calculated using a provincially approved method in accordance with the Provincial and National Management Information Systems guidelines.39,40 Resource use (ie, number of medications, laboratory tests, or consumables used; nursing care hours allocated per patient based on a recorded WLM) was measured for all patients.30 The number of units used by each patient then was multiplied by the cost per unit, which is estimated annually in the CHR,
LEE ET AL
to give an estimate of direct costs. Direct costs then were combined with an estimate of indirect costs (defined as costs not directly associated with patient care, ie, cost of nursing managers, administration) for every in-patient encounter. The quality of data reporting of costs in Alberta has been highly ranked by the Canadian Institutes for Health Information.41 Because of the limited number of patients treated with CCPD, patients treated with CAPD and CCPD were pooled together to give a more stable estimate of inpatient costs. Inpatient costs were excluded from all other cost categories to avoid double counting. Outpatient Nondialysis Expenses Outpatient nondialysis expenses included clinic visits, emergency room, day surgery, laboratory tests, radiology, and medications. The cost of outpatient care was determined by examining actual patient resource consumption and allocated nursing contact for patients treated with in-center, satellite, home/self-care hemodialysis, and peritoneal dialysis. The SARP database30 records the number of clinic visits made by each patient; cost per visit was determined by dividing the annual clinic staff’s wages (including wages of clinic nurses, reception staff, and nursing managers) by total number of clinic visits annually for all patients and adding a 20% overhead charge. The number of outpatient visits to the emergency room, day surgery unit, and radiology was determined from the CHR corporate database. The cost for each visit (including indirect costs and 20% overhead) was estimated based on average cost for a representative sample of CHR patients who visited the emergency room (categorized by reason for assessment) or received treatment in day surgery or radiology (categorized by procedure). The frequency with which laboratory tests were performed was recorded for each patient by the SARP database.30 This frequency was multiplied by unit costs assigned to each laboratory test based on a defined WLM. Medications were divided into three categories: erythropoietin (Eprex; Ortho Biotech, Toronto, Ontario, Canada), intravenous iron (Venofer), and others. Frequency and dosage were extracted from the SARP database for each patient.30 For erythropoietin and intravenous iron,
THE COST OF DIALYSIS
we multiplied the dosage by unit costs (CAN $18.45 [US $12.71] per 1,000 units for Eprex and CAN $30 (US $20.67) per 100 mg for Venofer) to determine cost per patient. Locally, all patients administered erythropoietin do so through the subcutaneous route. All other medication costs were determined from the Alberta Pharmaceutical Price List, summed to give a total cost per patient. In Alberta, the cost of all medications for patients with ESRD is covered through a government-sponsored insurance plan (except for a $25/mon maximum copayment). Physician Fees To maintain patient confidentiality, physician claims were acquired from Alberta Health and Wellness in aggregate groups of 6 to 10 patients, according to dialysis modality and comorbidity. Physician claims included visits to general practitioner, nephrologist, radiologist, and other subspecialists. REFERENCES 1. US Renal Data System: Economic costs of ESRD. Am J Kidney Dis 36:S163-S176, 2000 (suppl 2) 2. Garella S: The costs of dialysis in the USA. Nephrol Dial Transplant 12:10-21, 1997 3. Mallick NP: The costs of renal services in Britain. Nephrol Dial Transplant 12:25-28, 1997 4. Tomson CR: Recent advances: Nephrology. BMJ 320: 98-101, 2000 5. US Renal Data System: Incidence and prevalence of ESRD. Am J Kidney Dis 30:S40-S53, 1997 (suppl 1) 6. Goeree R, Manalich J, Grootendorst P, Beecroft ML, Churchill DN: Cost analysis of dialysis treatments for endstage renal disease (ESRD). Clin Invest Med 18:455-464, 1995 7. Wolfe RA, Ashby VB, Milford EL, et al: Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med 341:1725-1730, 1999 8. Laupacis A, Keown P, Pus N, et al: A study of the quality of life and cost-utility of renal transplantation. Kidney Int 50:235-242, 1996 9. Bloembergen WE, Port FK, Mauger EA, Wolfe RA: A comparison of mortality between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 6:177183, 1995 10. Collins AJ, Hao W, Xia H, et al: Mortality risks of peritoneal dialysis and hemodialysis. Am J Kidney Dis 34:1065-1074, 1999 11. Fenton SS, Schaubel DE, Desmeules M, et al: Hemodialysis versus peritoneal dialysis: A comparison of adjusted mortality rates. Am J Kidney Dis 30:334-342, 1997 12. Foley RN, Parfrey PS, Harnett JD, et al: Mode of dialysis therapy and mortality in end-stage renal disease. J Am Soc Nephrol 9:267-276, 1998
621
13. Merkus MP, Jager KJ, Dekker FW, Boeschoten EW, Stevens P, Krediet RT: Quality of life in patients on chronic dialysis: Self-assessment 3 months after the start of treatment. The Necosad Study Group. Am J Kidney Dis 29:584592, 1997 14. Churchill DN, Torrance GW, Taylor DW, et al: Measurement of quality of life in end-stage renal disease: The time trade-off approach. Clin Invest Med 10:14-20, 1987 15. de Wit GA, Ramsteijn PG, de Charro FT: Economic evaluation of end stage renal disease treatment. Health Policy 44:215-232, 1998 16. Guernsey L: Self-care for the ESRD patient. Nephrol Nurse 4:8-12, 1982 17. Sexton Dobby A: Self-care hemodialysis training and anatomy: The patient’s perspective. CANNT J 2(4):17-18, 1992 18. Prichard SS: Treatment modality selection in 150 consecutive patients starting ESRD therapy. Perit Dial Int 16:69-72, 1996 19. Churchill DN, Lemon BC, Torrance GW: A costeffectiveness analysis of continuous ambulatory peritoneal dialysis and hospital hemodialysis. Med Decis Making 4:489500, 1984 20. Garner TI, Dardis R: Cost-effectiveness analysis of end-stage renal disease treatments. Med Care 25:25-34, 1987 21. Tousignant P, Guttmann RD, Hollomby DJ: Transplantation and home hemodialysis: Their cost-effectiveness. J Chronic Dis 38:589-601, 1985 22. Ludbrook A: A cost-effectiveness analysis of the treatment of chronic renal failure. Appl Econ 13:337-350, 1981 23. Thompson K: Cost comparison of in-center dialysis, home dialysis, and transplantation in the Nashville area. Nephrology Nurse 2(6):24-25, 1980 24. Weydevelt FC, Bacquaert-Dufour K, Benevent D, et al: A cost-effectiveness analysis of continuous ambulatory peritoneal dialysis vs self-care in-center hemodialysis in France. Dial Transplant 28:70-74, 1999 25. Croxson BE, Ashton T: A cost effectiveness analysis of the treatment of end stage renal failure. N Z Med J 103:171-174, 1990 26. Deziel C, Hirsch DJ, Hoult P, et al: Clinical practice guidelines for the delivery of hemodialysis. Canadian Society of Nephrology. J Am Soc Nephrol 10:S306-S310, 1999 (suppl 13) 27. Churchill D, Taylor DW, Keshaviah P: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7:198207, 1996 28. Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 40:373-383, 1987 29. Beddhu S, Bruns FJ, Saul M, Seddon P, Zeidel ML: A simple comorbidity scale predicts clinical outcomes and costs in dialysis patients. Am J Med 108:609-613, 2000 30. Manns BJ, Mortis G, Taub K, McLaughlin K, Donaldson C, Ghali WA: The Southern Alberta Renal Program Database: A prototype for patient management and research initiatives. Clin Invest Med 24:164-170, 2001
622
31. MacLeod A, Grant A, Donaldson C, et al: Effectiveness and efficiency of methods of dialysis therapy for end-stage renal disease: Systematic reviews. Health Technol Assess 2:1-166, 1998 32. Hirth RA, Held PJ, Orzol SM, Dor A: Practice patterns, case mix, Medicare payment policy, and dialysis facility costs. Health Serv Res 33:1567-1592, 1999 33. US Renal Data System: 2001 Atlas of ESRD in the United States, in (vol 2002), United States Renal Data System, 2001 34. Nassar GM, Ayus JC: Infectious complications of the hemodialysis access. Kidney Int 60:1-13, 2001 35. National Kidney Foundation: K/DOQI: Clinical Practice Guidelines for Vascular Access: Update 2000. Am J Kidney Dis 37:S137-S181, 2001 (suppl 1) 36. Schwab SJ, Oliver MJ, Suhocki P, McCann R: Hemodialysis arteriovenous access: Detection of stenosis and response to treatment by vascular access blood flow. Kidney Int 59:358-362, 2001
LEE ET AL
37. World Bank: World Development Indicators 2000. Washington, DC, World Bank, 2001 38. Feachem RG, Sekhri NK, White KL: Getting more for their dollar: A comparison of the NHS with California’s Kaiser Permanente. BMJ 324:135-141, 2002 39. Canadian Institute for Health Information: Guidelines for Management Information Systems in Canadian Health Service Organizations. Ottawa, Canada, Canadian Institute for Health Information, 1999 40. McKillop I: A Research Project to Examine the Costing Methodologies Recommended in the MIS Guidelines. Ottawa, Canada, Canadian Institutes for Health Information, 1995, p 155 41. McKillop I, Pink GH, Johnson LM: The Financial Management of Acute Care in Canada: A Review of Funding, Performance Monitoring and Reporting Practices. Ottawa, Canada, Canadian Institute for Health Information, 2001, p 83